Systems and methods for power theft detection
View Patent ↗Systems, apparatuses, methods, and computer program products are disclosed for power theft detection. An example method includes receiving, by a control system, telemetry data from a transformer adjacent to a customer premise and a meter at the customer premise and storing, by the control system, the telemetry data in a memory. The example method further includes calculating, by the control system and using the telemetry data, a change in impedance in an electric line segment between the transformer and the meter, and determining, by the control system, whether the change in the impedance in the electric line segment is anomalous. Corresponding apparatuses and computer program products are also disclosed.
1. A method for power theft detection, the method comprising:
receiving, by a control system, telemetry data from a transformer adjacent to a customer premise and a meter at the customer premise;
storing, by the control system, the telemetry data in a memory;
calculating, by the control system and using the telemetry data, a change in impedance in an electric line segment between the transformer and the meter; and
determining, by the control system, whether the change in the impedance in the electric line segment is anomalous, wherein determining whether the change in the impedance in the electric line segment is anomalous includes:
retrieving, by the control system and from the memory, a plurality of previously calculated changes in the impedance in the electric line segment; and
determining, using a machine learning model and the plurality of previously calculated changes in the impedance in the electric line segment, whether the change in the impedance in the electric line segment is anomalous.
2. The method of claim 1 , further comprising:
in an instance in which the control system determines that the change the impedance in the electric line segment is anomalous, causing, by the control system, transmission of an alert indicating possible power theft.
3. The method of claim 1 , wherein the telemetry data is received via a fiber optic network.
4. The method of claim 3 , wherein the telemetry data is received via passive-optical networking.
5. The method of claim 1 , wherein the control system periodically receives the telemetry data from the transformer and the meter.
6. The method of claim 5 , wherein the control system periodically receives the telemetry data from the transformer and the meter at sub-second intervals.
7. The method of claim 1 , wherein calculating the change in the impedance in the electric line segment between the transformer and the meter includes:
retrieving, by the control system, impedance measurements from the transformer and the meter; and
calculating, by the control system, a difference between the impedance measurements from the transformer and the meter.
8. The method of claim 1 , further comprising:
training, by the control system, the machine learning model using a historical training data set comprising data regarding historical changes in impedance in electric line segments between transformers adjacent to customer premises and meters at the customer premises.
9. The method of claim 1 , wherein the machine learning model comprises a convolutional neural network.
10. An apparatus for power theft detection, the apparatus comprising a processor and a memory storing software instructions that, when executed by the processor, cause the apparatus to:
receive telemetry data from a transformer adjacent to a customer premise and a meter at the customer premise;
store the telemetry data in a memory;
calculate, using the telemetry data, a change in impedance in an electric line segment between the transformer and the meter; and
determine whether the change in the impedance in the electric line segment is anomalous, wherein determination of whether the change in the impedance in the electric line segment is anomalous includes:
retrieval of a plurality of previously calculated changes in the impedance in the electric line segment; and
determination of whether the change in the impedance in the electric line segment is anomalous with a convolutional neural network and the plurality of previously calculated changes in the impedance in the electric line segment.
11. The apparatus of claim 10 , the processor and a memory storing software instructions, when executed by the processor, further cause the apparatus to:
in an instance in which the change the impedance in the electric line segment is determined to be anomalous, cause transmission of an alert indicating possible power theft.
12. The apparatus of claim 10 , wherein the telemetry data is received via a fiber optic network.
13. The apparatus of claim 12 , wherein the telemetry data is received via passive-optical networking.
14. The apparatus of claim 10 , wherein the telemetry data is received periodically from the transformer and the meter.
15. The apparatus of claim 14 , wherein the telemetry data is received periodically from the transformer and the meter at sub-second intervals.
16. The apparatus of claim 10 , the processor and a memory storing software instructions, when executed by the processor and when calculating the change in the impedance in the electric line segment between the transformer and the meter, further cause the apparatus to:
retrieve impedance measurements from the transformer and the meter; and
calculate a difference between the impedance measurements from the transformer and the meter.
17. The apparatus of claim 10 , the processor and a memory storing software instructions, when executed by the processor, further cause the apparatus to:
train the machine learning model using a historical training data set comprising data regarding historical changes in impedance in electric line segments between transformers adjacent to customer premises and meters at the customer premises.
18. The apparatus of claim 10 , wherein the machine learning model comprises a convolutional neural network.
19. A computer program product for power theft detection, the computer program product comprising at least one non-transitory computer-readable storage medium storing software instructions that, when executed by an apparatus, cause the apparatus to:
receive telemetry data from a transformer adjacent to a customer premise and a meter at the customer premise;
store the telemetry data in a memory;
calculate, using the telemetry data, a change in impedance in an electric line segment between the transformer and the meter; and
determine whether the change in the impedance in the electric line segment is anomalous, wherein determination of whether the change in the impedance in the electric line segment is anomalous includes:
retrieval of a plurality of previously calculated changes in the impedance in the electric line segment; and
determination of whether the change in the impedance in the electric line segment is anomalous with a machine learning model and the plurality of previously calculated changes in the impedance in the electric line segment.
20. The computer program product of claim 19 , the at least one non-transitory computer-readable storage medium storing software instructions that, when executed by an apparatus, further cause the apparatus to:
in an instance in which the change the impedance in the electric line segment is determined to be anomalous, cause transmission of an alert indicating possible power theft.